Skip to content
Book a demo
Menu

Learn · glossary

The words, and what they mean here.

Planning has a vocabulary that is mostly stable across companies and completely unstable inside one. This page says what each term means as we use it, and where our own modelling words sit against the ones you already have. The acronyms stay in English in every language on this site, because that is how the reader’s own tools write them.

Last updated · 2026-09-18 Three groups · processes, planning, the model

01
The processes

The cycles a company runs, and what each one is for.

S&OP
Sales and operations planning. The monthly cycle in which demand, supply, finance and the executive team agree one set of numbers — how much will be sold, whether it can be made and moved, and what that does to the P&L. Its output is a consensus plan that everyone is then held to, which is why the argument is worth having in one model rather than four decks. S&OP in Superiora Planner.
FP&A
Financial planning and analysis. The finance function’s own cycle: the budget, the reforecast, the variance pack, and answering why a line moved. Its distinguishing feature is that it is measured against a commitment — the plan of record — rather than against the last version of itself. FP&A in Superiora Planner.
xP&A
Extended planning and analysis. FP&A stretched past finance into the operational plans it depends on — workforce, supply, sales capacity, marketing — so that a change in one is visible in the others without a reconciliation meeting. It is less a new process than the admission that a finance plan built on numbers nobody operational recognises is not a plan. xP&A in Superiora Planner.
DRP
Distribution requirements planning. Working out what stock has to be where, and when, across a network of warehouses and depots: demand at each location, the cover it needs, the lead time to replenish it, and the flows between locations that follow. It answers “where does it go” rather than “how much do we make”. DRP in Superiora Planner.
MRP
Material requirements planning. Exploding a production plan through the bill of materials into the components and raw materials it consumes, offset by their lead times, and netting that against what is already in stock or on order. It answers “what must we buy, and by when”. MRP in Superiora Planner.
Demand forecasting
Estimating what will be sold, usually by fitting a statistical baseline to history and then adjusting it with what the business knows that the history does not — a promotion, a listing gained, a competitor leaving. The forecast is the input to almost everything else, which is why its errors are expensive. Demand forecasting in Superiora Planner.

02
Planning terms

The words that get used loosely, defined tightly.

Decision
A choice a person or a committee makes, in a window, that fixes what the plan underneath it is allowed to say — a range signed off, an envelope given to a function, a plant opened. In Superiora a decision is an object in the model with an owner, a date and its reasoning, not a note in a deck beside it: what it reaches can be computed, so changing it recalculates everything under it.
Driver
The quantity a plan is built from rather than typed into — a price, a conversion rate, a cost per unit, a productivity assumption. Driver-based planning means the figures are derived from a small set of these instead of being entered line by line, which makes a variance explainable in one line rather than in three meetings.
Plan of record
The approved, dated, frozen version everything else is measured against. Re-baseline it casually and variance stops meaning anything; refuse ever to re-baseline and by the third quarter nobody looks at it. Either way it is a decision, and it is better taken in the open.
Switchover
The boundary between actual and forecast: the last closed period. Everything to the left of it is what happened; everything to the right is what is planned. One setting moves it for the whole model, and moving it is the act that turns a forecast month into a closed one.
Period lock
A closed period that refuses every write into it — typed, pasted, imported, scheduled, or made by the system itself. It is what makes a pack issued in March still reconcile in November, and what makes last year’s comparison stable enough to argue with.
Scenario
A named alternative held beside the working numbers: the same model with one or more decisions or drivers changed, so the consequence is derived rather than retyped. A scenario is an argument you can read next to the base, not a copy of the file with a different name.
Version
A stage of the same plan — Actual, Budget, Forecast — held as its own axis of the model, so the P&L can be read in any of them and variance between two of them is a subtraction rather than a reconciliation.
Rolling forecast
A forecast that always looks the same distance ahead: actuals to date, plus a plan for the remaining periods, re-cut each cycle. It is affordable only when the reforecast rule says what may move — otherwise it is a monthly budget round with a friendlier name.

03
The model

What the platform calls the parts of a model.

If you have built in Anaplan, the semantics will be familiar and several of the names are the same. Where ours differ, both are given.

Dimension
An axis of the model and the list of members on it: products, stores, cost centres, months. Members can be hierarchical, so a parent aggregates its children — sometimes with a weight, which is how a contra or expense line rolls up negatively. (An Anaplan modeller would call this a list, and the platform’s API answers to both names.)
Hypercube
A grid over a set of dimensions — Sales by Store by SKU by Week, say — holding the measures that share that grain. Choosing the dimensions of a cube is the first modelling decision and the expensive one to change. (Anaplan: a module.)
Measure
A line in a cube. Either an input, which a person or an import writes into, or a formula, which is computed and cannot be typed over. A formula measure is read-only by construction, not by convention. (Anaplan: a line item.)
Formula
The expression that defines a calculated measure, written against other measures by name with the dimensions aligned automatically. It is checked against the shape of the model when it is saved — so a dimensional mismatch or a circular reference is refused, by name, while you are still thinking about it rather than when the model is run.
Selective access
Also dimensional access. Granting a user specific slices of a model — a region, an entity, a set of cost centres — rather than the whole of it. Grants are explicit, and they are enforced on the read side, so a cell you may not see is never offered for writing.
Changeset
A set of edits staged together and applied as one revision, with its impact visible before it lands. If it was built against a revision that has since moved, it conflicts as a whole rather than half-landing — there is no silent last-write-wins.
Audit log
The append-only record of every change to a model: what changed, by whom, when. State is rebuilt by replaying it rather than edited in place, which is why a restart returns the model exactly as it was and why a figure can always be walked back to the act that produced it.
Revision
A published state of the model, produced by applying a changeset. Readers see a consistent revision while a write is in progress, so a report never catches the model halfway through a change.

Bring your decisions. Superiora does the planning.

We stand the decision model up on your own goals, data and KPIs, then build the plans and stress the scenarios underneath them. An hour to see it on your own numbers.

A form, or your mail client — both reach the founding team